Abstract
BACKGROUND/OBJECTIVES: Vitamin D's effect on risk health outcomes is often evaluated using prospective cohort studies. For vitamin D, risk ratios (RRs) are based on health outcomes with respect to serum 25-hydroxyvitamin D [25(OH)D] concentrations measured at time of enrollment. Serum 25(OH)D concentrations vary over time, thereby diluting the effect of 25(OH)D for long follow-up periods. Inverse relationships between RR and follow-up period have been reported for all-cause mortality rate and cancer incidence rates. Here, the effect for neurological outcomes is evaluated. METHODS: The analysis examines how follow-up period affected results from nine cohort studies of all-cause dementia, six studies of Alzheimer's disease, and nine for cognitive impairment with respect to vitamin D deficiency. RESULTS: For all-cause dementia, Alzheimer's disease, and cognitive impairment, respectively, the linear regression fits are RR = 2.9 - 0.14 × years, r = 0.73, p = 0.02; RR = 2.9 - 0.14 × years, r = 0.69, p = 0.13; and RR = 1.8 - 0.066 × years, r = 0.72, p = 0.03. The regression fit to RR for the shortest follow-up period for each outcome is considered the best estimate of vitamin D deficiency's effect on risk. Those values are approximately twice that found by averaging all RRs without considering the effect of follow-up period. CONCLUSIONS: Vitamin D's effect on risk of neurological conditions is inversely correlated with mean follow-up period in prospective cohort studies. This effect should be considered in the design and analysis of such studies. Additional studies should also be conducted regarding raising serum 25(OH)D concentrations to reduce risk of brain function decline.